three theories of the effects of language education programs: …macswan/macswanetal2017.pdf ·...

Annual Review of Applied Linguistics, 37 (2017), pp. 218–240.© Cambridge University Press, 2017doi: 10.1017/S0267190517000137

Three Theories of the Effects of Language EducationPrograms: An Empirical Evaluation of Bilingual and

English-Only Policies

Jeff MacSwan

University of Maryland

Marilyn S. Thompson

Arizona State University

Kellie Rolstad

University of Maryland

Kara McAlister


Gerda Lobo


abstract

We empirically evaluated three theoretical models—the threshold hypothesis, trans-fer theory, and time-on-task theory—for educating English language learners (ELLs),with a focus on the role of language factors in explaining achievement differencesamong ELLs. Participants were 196 sixth graders with Spanish language backgroundswho started learning English in kindergarten and then were continuously enrolled ina U.S. school. Structural equation modeling was used to estimate the extent to whichSpanish and English language and literacy—skills that are emphasized differentiallyin competing theories for educating ELLs—predict academic achievement assessedin English. Results indicated that Spanish literacy, over and above English languageproficiency, was substantially predictive of academic achievement, consistent withthe transfer theory. This model was a more focused version of the threshold hypoth-esis, in that the weaker predictor of Spanish oral language proficiency was excluded.Time-on task theory was not supported.

introduction

Young second language (L2) learners in U.S. schools, or bilingual learners, repre-sent a substantial proportion of the school-age population. Known in policy con-texts as English language learners (ELLs), these students represented 9.3% of allU.S. public school children in 2013–2014. Between 6% and 9.9% of students wereELLs in Arizona, Arkansas, Delaware, Florida, Hawaii, Illinois, Kansas, Mary-land, Massachusetts, Minnesota, New York, North Carolina, Oklahoma, Oregon,

218

Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/S0267190517000137Downloaded from https://www.cambridge.org/core. University of Maryland - University Libraries, on 27 Sep 2017 at 02:57:18, subject to the

http://dx.doi.org/10.1017/S0267190517000137

http://crossmark.crossref.org/dialog/?doi=10.1017/S0267190517000137&domain=pdf

https://www.cambridge.org/core/terms

https://doi.org/10.1017/S0267190517000137

https://www.cambridge.org/core

three theories of the effects of language education 219

Rhode Island, Virginia, and Washington; whereas Alaska, California, Colorado,Nevada, New Mexico, and Texas had ELL enrollments of 10% or more. California’sELL population was the largest, at 22.7%. Most of these students (76.5%) spokeSpanish as a first language. The National Assessment of Educational Progress hasdocumented a substantial achievement gap between ELLs and non-ELLs sincedata collection began in the 1970s, and this gap persists to the present day. Thisachievement gap underscores the special challenges ELLs face in school. (Formore details. see Kena et al., 2016.)

In a landmark decision, Lau v. Nichols (1974), the Supreme Court unanimouslyfound that a “disparate impact” occurred when a San Francisco school districtfailed to provide for the special challenges facing school-age L2 learners, as“students who do not know English are effectively foreclosed from any meaningfuleducation.” The Court found that the policy violated the Civil Rights Act of 1964,and that the school district must provide students with “appropriate relief.”School districts have responded differently to Lau as well as other important courtdecisions and legislation. Some have implemented intensive programs to teachEnglish and maximize children’s exposure to English, such as structured Englishimmersion (SEI), while others use children’s first language (L1) as a support tohelp them keep up academically in school subjects while learning English.

A number of studies have been carried out to evaluate the relative effectivenessof different educational programs for bilingual learners. These studies haveestablished a strong consensus in the field favoring L1 support, a programapproach known as bilingual education, but relatively little empirical researchhas been done to evaluate specific theories underlying the positive effects ofL1 support. Moreover, English-only approaches to educating ELLs persist inmost states.1 In this article, we seek to move beyond the program effectivenessliterature to identify the extent to which first and second language and literacy—emphasized differently in various approaches to educating ELLs—explainachievement differences among ELLs. In particular, we employ a structuralequation modeling (SEM) approach to empirically evaluate three theoreticalmodels proposed to explain academic achievement differences among ELLs—thethreshold hypothesis, transfer theory, and the time-on-task theory—with a focuson the role language plays in such theories.

research on language minority education

Program Effectiveness Research: Description

Slavin and Cheung (2005) conducted a “best evidence” synthesis of researchon bilingual education programs focusing on methods of teaching reading toELL students, using methods typical of narrative reviews. They comparedacademic outcomes between ELL students taught using a bilingual approach(with the L1 used before or simultaneously with the L2 in reading instruction) andthose taught using an SEI approach (with reading instruction in English only).Slavin and Cheung’s review, based on a corpus of 16 studies, concluded that most



https://doi.org/10.1017/S0267190517000137


220 jeff macswan et al.

methodologically rigorous studies favored bilingual approaches over English-onlyapproaches; although some did not find a difference, none favored English-onlyapproaches over bilingual approaches. Thus, Slavin and Cheung concluded thatthe available evidence favored bilingual approaches, especially paired bilingualstrategies that teach reading in the native language and English at the same time.

While most other narrative reviews (August & Hakuta, 1997; August & Shana-han, 2006; Fillmore & Valadez, 1986; Genesee, Lindholm-Leary, Saunders, &Christian, 2006; Meyer & Fienberg, 1992; National Academies of Sciences, En-gineering, and Medicine, 2017) have reached similar conclusions, two reviewsco-authored by Keith Baker (Baker & de Kanter, 1981; Rossell & Baker, 1996)were more critical of bilingual education and remain highly influential amongadvocates of English-intensive programs. While the Baker reviews are often citedas favoring English-only approaches, the authors’ own conclusions state only thatpolicy should not promote an exclusive reliance on bilingual education, but shouldsupport the use of an SEI approach as well (Baker & de Kanter, 1981, chap. 4, p. 6;Rossell & Baker, 1996, p. 19). The Baker reviews have been criticized for includingstudies that did not meet the authors’ stated selection criteria, for assigning multi-ple “votes” to studies published in different forms, and for misclassifying programmodels (Greene, 1998; Krashen, 1996; McField & McField, 2014; Rolstad, Ma-honey, & Glass, 2005; Reljic, Ferring, & Martin, 2015; Slavin & Cheung, 2005).

In addition to narrative reviews, several authors have also conducted meta-analyses of the effectiveness of language education programs for ELLs. For in-stance, Willig (1985), Greene (1998), and Rolstad et al. (2005) found bilingualeducation to be superior to English-only approaches, and Rolstad et al. addition-ally discovered that developmental bilingual education programs, which continueto support children’s L1 long after they develop proficiency in English, wereeven more effective at promoting children’s academic achievement than transi-tional, early-exit bilingual education programs, which emphasize transitioning toall-English instruction as soon as possible. Rolstad et al. reported that studiescontrolling for ELL status indicated a positive effect for bilingual education of0.23 standard deviation, with outcome measures in the native language showinga positive effect of 0.86 standard deviation. More recently, McField and McField(2014) conducted a meta-analysis of all previous meta-analyses of bilingual edu-cation in the United States, and Reljic et al. (2015) conducted a meta-analysis ofbilingual education programs in Europe, modeled after Rolstad et al., both drawingsimilar conclusions regarding the effectiveness of bilingual approaches. Overall,this extensive body of meta-analytic research on the effectiveness of bilingualeducation programs has consistently found the use of the home language at schoolto be more effective than English alone at promoting ELL children’s academicachievement.

Program Effectiveness Research: Theory

While considerable research has been done on the effectiveness of bilingual educa-tion programs, little empirical research has been done to evaluate specific theories



https://doi.org/10.1017/S0267190517000137



Type of Bilingualism Cognitive Effects

Additive bilingualism. High levels in both languages

Positive cognitive effects

Dominant bilingualism. Native-like level in one of the languages

Neither positive nor negative cognitive effects

Higher threshold level of bilingual competence

Semilingualism. Low level in both languages (may be balanced or dominant)

Negative cognitive effects

Lower threshold level of bilingual competence

Lev

el o

f B

iling

ualis

m O

btai

ned

figure 1. The Threshold Hypothesis (adapted from Cummins, 1979, p. 230).

about the underlying causes of the positive effects associated with the use ofchildren’s home language in instructional settings. In the present study, we empir-ically evaluate three such theories—the threshold hypothesis, transfer theory, andthe time-on-task theory.

The threshold hypothesis (Cummins, 1976, 1979, 2000) is a leading theory ofacademic achievement differences among bilingual students. It posits that languageminority children undergo native language loss, or a shift in language dominancefrom the L1 to the L2, and that “the level of linguistic competence attained by abilingual child may mediate the effects of his bilingual learning experiences oncognitive growth.” Cummins (1979, 1981) included both literacy and language inhis conception of “low ability in both languages.” While it will doubtless be truethat children with low English literacy will also have low academic achievement,essentially by definition, the more provocative claim that Cummins made is thatlinguistic minorities in the United States are characterized by “less than native-likecommand of the vocabulary and syntactic structures of both L1 and L2” (Cummins,1979, p. 238), and that variation in L1 oral language is indicative of “low ability” inthe L1 (Cummins, 1994, p. 3814). Cummins (1979, p. 230) presented the thresholdhypothesis graphically (Figure 1).

MacSwan (2000) noted a concern with the threshold hypothesis in that it positsthat some children may have limited (oral) proficiency in their L1 while profi-ciency in the L2 is still developing, a condition Cummins and others (Hansegård,1968; Ringbom, 1962; Skutnabb-Kangas, 1981; Toukomaa & Skutnabb-Kangas,1977) have called “semilingualism.” The existence of such a condition was never



https://doi.org/10.1017/S0267190517000137



established empirically, and a review of the evidence has shown that it does notoccur in typically developing bilingual children (Commins & Miramontes, 1989;Edelsky, 2006; Edelsky et al., 1983; Hakuta & D’Andrea, 1992; MacSwan, 2000;MacSwan & Rolstad, 2003, 2006, 2010; Martin-Jones & Romaine, 1986; Paulston,1983; Petrovic & Olmstead, 2001; Valadez, MacSwan, & Martínez, 2000; Wiley& Rolstad, 2014; Wiley, 2005).

Different facets of language proficiency are included in Cummins’s concep-tion of bilingualism as embedded in the threshold hypothesis. Our analysis willtherefore examine various aspects of language (use) referenced by Cummins, to-gether and separately, and attempt to determine what variability exists amonglearners and whether these variations in abilities relate to success at school, asthe threshold hypothesis posits. By unpacking language proficiency into discretecomponents of oral language and literacy, and evaluating the effects of each onacademic achievement, we hope to gain a more nuanced view of the hypothesizedcausal relationships illustrated in Figure 1.

As a complement to the threshold hypothesis, Cummins (1981, p. 29) alsotheorized about the advantage of bilingual education in terms of transfer theory inthe linguistic interdependence hypothesis:

To the extent that instruction in Lx [(Language x)] is effective in promoting profi-ciency in Lx, transfer of this proficiency to Ly [(Language y)] will occur providedthere is adequate exposure to Ly (either in school or in the environment) and adequatemotivation to learn Ly.

To illuminate, Cummins (1981) contrasted the perspective of a common un-derlying proficiency with a separate underlying proficiency model; the commonunderlying proficiency model, which Cummins endorsed, relates separate “surfacefeatures” of both languages to a common underlying proficiency which is sharedbetween them, facilitating transfer. As Cummins more recently elaborated,

In concrete terms, what this hypothesis implies is that in, for example, a dual languageSpanish-English bilingual program in the United States, Spanish instruction thatdevelops Spanish reading and writing skills is not just developing Spanish skills, it isalso developing a deeper conceptual and linguistic proficiency that is strongly relatedto the development of literacy in the majority language (English). In other words,although the surface aspects (e.g., pronunciation, fluency, etc.) of different languagescan be distinguished, there is an underlying cognitive/academic proficiency that iscommon across languages. This common underlying proficiency makes possible thetransfer of cognitive/academic or literacy-related proficiency from one language toanother. (Cummins, 2017, p. 106)

Krashen (1996) similarly emphasized the role of transfer across languages, butwithout specifically drawing upon the linguistic interdependence hypothesis. ForKrashen, bilingual instruction reduces achievement differences between bilingualand monolingual English-speaking children because it allows ELLs to keep upacademically while they struggle to learn and understand English; this idea was



https://doi.org/10.1017/S0267190517000137



the basic rationale for bilingual education programs noted by the United StatesCommission on Civil Rights (1975) and embedded in Lau v. Nichols (1974).Krashen argued that language minority children in all-English instruction whoachieve well are able to do so because they are likely to have educational resourcesat home that provide a kind of “de facto bilingual education”; that is, althoughthese children cannot understand academic instruction at school because they donot know English, they receive L1 support for school subject learning at home. Inaddition, Krashen argued, exposure to academic content in L1 allows children toachieve higher levels of proficiency in L2 because it provides a network of contex-tual clues for children to make inferences regarding the meaning of new words andlinguistic structures presented in English. There is substantial empirical evidencethat knowledge of school subjects, and literacy in particular, transfers across chil-dren’s two languages (for summaries, see August & Shanahan, 2006; Cummins,2017; Genesee et al., 2006; National Academies of Sciences, Engineering, andMedicine, 2017).

As MacSwan and Rolstad (2005) have noted, a concern with the linguistic inter-dependence hypothesis is the embedded assumption that language and language-related academic content matter are not distinct. The surface features of the L1and L2 for Cummins (1981, p. 25) are those that “have become relatively au-tomatized or less cognitively demanding, whereas the underlying proficiency isthat involved in cognitively demanding communicative tasks,” that is, in-schooltalk developed around academic tasks. If language proficiency is understood toinclude aspects of school knowledge, as in Cummins’s model, then it followsthat growth in academic subjects will result from instruction in the L1. Cumminsthus captures the transfer effect in his model of bilingualism, but the model isplagued by a problem: Pedagogical content knowledge is characterized as part ofour “underlying cognitive/academic proficiency that is common across languages”rather than as discretely represented, context-dependent conceptual knowledge. Assuch, the linguistic interdependence hypothesis implies that language developedaround nonschool contexts, typically used by lower socioeconomic communities(e.g., African American English and working-class English), is less developed thanthe language variety used at school. This aspect of the linguistic interdependencehypothesis makes the theory vulnerable to some of the same criticisms that havebeen levied against the threshold hypothesis (MacSwan, 2000).

MacSwan and Rolstad (2005) developed an approach to transfer theory thatspecifically differentiates between language and conceptual understanding ofschool subject matter, drawing on neurocognitive research around psychologicalmodularity. In this view, language is cognitively special and discrete; all typicallydeveloping children acquire the language of their speech community, effortlesslyand without instruction, shaped by the specific contexts in which it is acquiredas an “accidental product of varied experience” (Chomsky, 1995, p. 6). Languageclearly links to the conceptual knowledge developed in school, but knowledge ofschool subjects and other topics is distinct from linguistic knowledge itself andnot part of our linguistic development. (For summaries of relevant psycholinguisticresearch, see Curtiss, 2013; Gallistel & King, 2009.) Viewed in these terms, transfer



https://doi.org/10.1017/S0267190517000137



is simply a metaphor for the accessibility of conceptual knowledge through thevarious languages people may know. Because transfer occurs across languagesin this way, learning school subjects in an L1 develops conceptual knowledgeavailable to children as they develop proficiency in their L2.

This conception of the cognitive underpinnings of transfer is consistent with theview that multilingual speakers have an underlying integrated language system,with both shared and discrete linguistic resources (MacSwan, 2017). For instruc-tional purposes, languages might be used separately or together, using codeswitch-ing, translation, or other syncretic processes, depending on a teacher’s objectives,as many have suggested (Cook, 2001; Cummins, 2008, 2017; Faltis, 1989; García,2009; Jacobson, 1978, 1990; Milk, 1986). Content knowledge is not specificallyassociated with one language or the other, but is gained and accessed through either.

The study reported here specifically evaluates this version of transfer theoryrather than the similar but somewhat differently conceptualized linguistic interde-pendence hypothesis.

Finally, Rossell and Baker (1996) proposed that “time-on-task” in L2 is thecentral factor underlying achievement among ELLs. According to their time-on-task theory, the more time children spend hearing, speaking, or studying Englishthe higher their level of L2 proficiency will be, which in turn will translate intohigher scores on academic achievement measures. As Rossell and Baker (1996,p. 22) put it, “The ‘time-on-task’ principle [is] … the notion that the amount of timespent learning a subject is the greatest predictor of achievement in that subject.”Porter (1990), also an advocate of English-only instruction, defined the time-on-task principle in this way: “The more time spent learning a language, the betteryou do in it, all other factors being equal” (p. 119). Like the linguistic interdepen-dence hypothesis, the time-on-task theory does not distinguish between learninga language and learning school content in a language. The difference is that thelinguistic interdependence hypothesis posits the existence of a common underlyingproficiency as a repository of some aspects of school content knowledge, whereasthe time-on-task theory does not.

Each of the theories described recommends a treatment aimed at achieving aspecific intermediate effect, which in turn is hypothesized to predict academicachievement outcomes in English. The threshold hypothesis and transfer theoryboth recommend bilingual education as an instructional strategy. For the thresholdhypothesis, the hypothesized intermediate cognitive effect is increased overallbilingualism (understood to be language proficiency, inclusive of both literacyand oral language in both L1 and L2), with the predicted result that participantswill achieve higher academic achievement in English. The transfer theory, on theother hand, hypothesizes that bilingual education will have an intermediate effectof increasing academic achievement in the L1 as participants concurrently learnEnglish, with the predicted result that children will transfer this knowledge to theL2 environment. The time-on-task theory, by contrast, recommends that childrenspend as much time as possible in English, with the intermediate effect of increasedEnglish language proficiency and predicted result of higher academic achievementin English. These relationships are summarized in Table 1.



https://doi.org/10.1017/S0267190517000137



table 1. Structural Components of Theories of Academic Achievement DifferencesAmong English Language Learners

TheoryAssociated

Program Intermediate Effect Predicted Outcome

Threshold Hypothesis BilingualEducation

Improved overallbilingualism (languageand literacy)

Higher academicachievement measuredin the L2

Transfer Theory BilingualEducation

Improved academicachievement throughL1 instruction


Time-on-Task Theory Structured EnglishImmersion

Increased proficiency inthe L2


We may thus think of the transfer theory as a more focused version of thethreshold hypothesis. Transfer theory looks to improved L1 literacy, a componentof academic achievement that is developed concurrently with English languageproficiency, as an intermediate effect. The threshold hypothesis, on the other hand,posits that both L1 oral language and L1 literacy are related to increased academicachievement in the L2.

The underlying causes of student success are clearly much more complex thanany of these models propose. The degree to which the recommended treatment islikely to achieve the intermediate effect will no doubt be moderated by numerousfactors besides language of instruction, such as teacher quality, school resources,instructional methods, and more (August & Hakuta, 1997; Genesee et al., 2006;National Academies of Sciences, Engineering, and Medicine, 2017). However,as previously noted, there are clear overall effects stemming from language ofinstruction; here we focus narrowly on this effect, evaluating the link between ahypothesized intermediate effect and the predicted outcome to inform the meritsand utility of each of these three theories as explanatory models. For Spanishacademic achievement measures, we focus on Spanish literacy.

methods

Research Questions

We empirically examine the merits of each of the three theories by addressing thefollowing research questions:

1. Threshold hypothesis: Do both Spanish language proficiency (Spanish language andliteracy) and English language proficiency (English language and literacy) predictacademic achievement measured in English?

2. Transfer theory: Does Spanish literacy, in particular, contribute over and aboveEnglish language proficiency to the prediction of academic achievement measuredin English?



https://doi.org/10.1017/S0267190517000137



3. Time-on-task theory: Is English language proficiency sufficient as a predictor ofacademic achievement measured in English, reflecting English-only approaches ofthe time-on-task theory, or is prediction of achievement improved by inclusion ofSpanish language and Spanish literacy?

Participants

Study participants came from seven school districts in a major metropolitan area.To be included in the sample, participants had to be Spanish language-backgroundsixth graders at the time of the study who had been continuously enrolled in a U.S.school since kindergarten, and who had not started learning English until kinder-garten. Participants were educated in a variety of minority language educationalprograms, including English as a second language (ESL), bilingual education, duallanguage immersion, and SEI.2 For data collection, each year over a 3-year period,we obtained a list of all ELLs and reclassified ELLs entering the sixth grade in allparticipating school districts. We sent letters of invitation home; the letters includeda brief survey to determine whether the participants met the selection criteria. Ofthose who responded, 196 children were identified as eligible to participate in thestudy.

The assessments noted in the next section were administered in clusters foreach language, with at least 2 weeks between the matching Spanish and Englishversions of tests to avoid priming effects.

Instrumentation

Language and Literacy Measures. Using standard methods in the study of childlanguage, students were asked to interact with native speakers of English andSpanish on separate occasions spanning 4–6 weeks and tell a story about a boyand a frog from Mercer Mayer’s picture book with no text (Mayer, 1969). Thesespeech samples were coded for lexical, morphological, and syntactic structuresand errors.

Speech samples of each child telling the whole story depicted in the picturebook were individually videotaped. These speech samples were then transcribedword for word and coded using MacWhinney’s (1995) standard CHAT format, asmodified by Curtiss, MacSwan, Schaeffer, Kural, and Sano (2004) and adaptedto Spanish by Valadez et al. (2000). An example of a coded Spanish utterance ispresented next, with translation provided in brackets:

∗MAR: El niño se está durmiendo, y la rana se escapó.[The boy is going to sleep, and the frog escaped.]

%mor: DART|el niño REF|se IAUX|está-3Ss dormir-DUR conj|y DART|la rana REF|seIF|escapar-pret-3Ss

%lex: N|niño N|dormir N|rana N|escapar

All transcripts were coded by native speakers, and each coded transcript wasproofed by at least one other native speaker trained in the coding system. Whenjudgments regarding error coding differed, these were resolved using procedures



https://doi.org/10.1017/S0267190517000137



detailed in Curtiss et al. (2004). In this coding system, errors of selection (e.g.,when la is used where el is required for the morphological category DART) wereprefixed with = (equal sign); errors of omission (a category such as DART orIAUX is missing altogether) were suffixed with =0. Erroneous lexical selectionwas similarly noted on the %lex: (lexical) tier, and errors in word order, if present,were noted with appropriate annotations on a %syn: (syntactic) tier. This systempermitted the calculation of morphological and syntactic error rates in English andSpanish for the children in the study. Because morphological structure is regardedas central to grammatical representation, both historically in the study of child lan-guage acquisition (Brown, 1973) and in contemporary linguistic theory (Chomsky,1995; Uriagereka, 2012), we focused our analysis on morphological error rate as anindex of participants’ knowledge of grammatical structure in English and Spanish.

In addition, we administered the Woodcock Language Proficiency Battery(WLPB) in English and Spanish to assess oral language proficiency and literacy.Literacy measures included the Basic Reading Skills and the Reading Compre-hension Clusters of the WLPB in English and Spanish, which focus on passagecomprehension, vocabulary, phonics skills, and structural analysis (Woodcock,1991). These two composite scores together were preferred over the Broad Read-ing composite, as they assess a wide range of reading skills. (In addition, becauseBroad Reading shares subscales that are used in the Basic and Comprehensioncomposites, including the Broad Reading composite in models with the othercomposites would have been statistically inappropriate due to overlapping scalesacross the composite.) Productive measures of vocabulary were collected using theClinical Evaluation of Language Fundamentals (CELF, in English and Spanish;Semel, Wiig, & Secord, 1995) in English and Spanish, and receptive vocabularyknowledge was tested using the Peabody Picture Vocabulary Test (PPVT; English)and Test de Vocabulario en Imágenes Peabody (PVIP; Spanish).

Of the oral language measures, we regard the language samples to be verysolid measures of children’s knowledge of grammatical structure in English andSpanish. Although a standard stimulus was used to elicit the samples, the contextof the task provided a broad opportunity for participants to demonstrate theirknowledge of the grammatical structure of English and Spanish. The English OralLanguage Cluster of the WLPB, on the other hand, is focused on English usedin academic contexts and may measure language proficiency concurrently withaspects of academic achievement (MacSwan & Pray, 2005). Similarly, we notethat the Peabody appears similarly to focus on school-related vocabulary and mayhave substantial construct overlap with academic achievement, unlike the CELF,an open-ended and more context-independent vocabulary measure. We thereforeregard the natural language samples and CELF to be good indicators of languageproficiency and believe the Woodcock Oral Language measure and the Peabodyvocabulary test to be more narrowly representative of language used in the contextof schooling.

Language measures were collected in both Spanish and English for all childrenin the study. Key constructs and their measures are summarized in Table 2. In theanalyses, we initially consider oral language and literacy as separate constructs,



https://doi.org/10.1017/S0267190517000137



table 2. Operationalizations of Language-Related Constructs

Construct Measures in Both Versions: Spanish (L1) and English (L2)

Oral Language ProficiencyGrammar Language sample (frog story)Vocabulary Clinical Evaluation of Language Fundamentals (CELF)

Peabody (Spanish PVIP; English PPVT)Woodcock Oral Language Cluster

Literacy Woodcock Reading ComprehensionWoodcock Basic Reading Skills

with oral language including measures of grammatical structure as well as vocab-ulary.

Academic Achievement Measures. Academic achievement was measured usingthe Reading, Mathematics, and Language scores of the Stanford Achievement Test,9th Edition (SAT-9), as administered at the end of the sixth grade and obtainedfrom students’ academic files. As is evident from the data summaries and analysespresented in the next sections, participants in the study appear to have had suf-ficiently well-developed English language proficiency to minimize measurementerror introduced by limited proficiency in the linguistic medium of the test (fordiscussion, see Thompson, DiCerbo, Mahoney, & MacSwan, 2002).

Analysis

A structural equation modeling (SEM) framework was used to examine the hy-pothesized relationships between English and Spanish language proficiency andacademic achievement as assessed in English. All analyses were conducted us-ing full-information maximum likelihood estimation within Mplus 6.11, whichaccommodates data missing at random. Preliminary analyses were conducted toexamine measures descriptively, including computations of means and standarddeviations for all English and Spanish language measures and for the academicachievement outcomes, as well as correlations among the measures.

The modeling process consisted first of an evaluation of measurement modelsrelating the hypothesized constructs to the language and achievement measures,followed by the estimation of latent regression models in which the constructs un-derlying English and Spanish language proficiency predict academic achievement.The global fit of each model was evaluated according to the model chi-squarestatistic and several goodness-of-fit indices, including the comparative fit index(CFI), root mean square error of approximation (RMSEA), and the standardizedroot mean square residual (SRMR). The Akaike information criterion (AIC) wasalso included as a parsimony-corrected index to support comparisons of fit acrossnonnested models; smaller values indicate better fit. The proportions of variancein the academic achievement factor explained by the latent regression models werealso examined using R2 indices.



https://doi.org/10.1017/S0267190517000137



Three alternative measurement models were first fit to the data. These mea-surement models differed in the number and nature of factors used to modelcovariances among the six proficiency measures within each language: (a) withseparate factors for literacy, school-based language, and natural language; (b)with separate factors for literacy and oral language development (i.e., com-bining school-based and natural language into a single factor); and (c) witha single language proficiency factor. It was expected that the language factorswould be strongly correlated within both Spanish and English, so although anuanced view of the structure of language was desired, we had concerns aboutencountering multicollinearity problems in the latent regression model in whichthese correlated factors predicted academic achievement. Finally, latent regres-sion models were then specified in which academic achievement was predictedfrom the English and Spanish language factors supported in the measurementmodels.

results

Preliminary Analyses

Based on an analysis of multivariate outliers, data for five students were omittedfrom further analyses due to high Mahalanobis distances, influence statistics, andCook’s distances. The resulting data set for analysis consisted of 191 cases. Per-centages of cases with missing data for the measures ranged from less than 2%for the Spanish Peabody to approximately 20% for SAT-9 achievement measures.The major reasons for missing observations were absence during administrationfor select English and Spanish language measures and, for reasons unknown andunavailable to us, missing SAT-9 scores in the school data files. Estimates of cor-relations, means, and standard deviations for the measures are reported in Table 3,as computed in the full-information maximum likelihood procedure used in allSEM analyses.

As expected, the language measures most strongly correlated with the threeacademic achievement measures (i.e., reading, language, mathematics) were theEnglish reading comprehension, basic reading, and oral/vocabulary language mea-sures; these correlations all exceeded .5 except for the one between mathematicsand English basic reading. Additionally, several measures of Spanish literacy andlanguage were substantially correlated with the English achievement outcomes.Correlations between Spanish reading comprehension and the three achievementmeasures ranged from .39 (mathematics) to .47 (reading); Spanish basic reading,oral/vocabulary, and the Peabody were also correlated with achievement outcomesin the .3 to .4 range. In contrast, the language sample measures (i.e., frog storyand CELF) were minimally correlated with the achievement outcomes, with theexception of the English CELF’s correlation with reading (.39) and mathematics(.44). Scores on the frog story tended to be particularly high and to display littlevariance, indicating students in the sample were proficient speakers of both lan-guages (rates of correct morphological responses: English M = 97.75, SD = 1.35;



https://doi.org/10.1017/S0267190517000137


230je

ffm

ac

swa

ne

ta

l.

table 3. Correlations and Descriptive Statistics Among Measures

Measures 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

1. Sp. Comp. 1.0002. Sp. Basic 0.657 1.0003. Sp. Peabody 0.480 0.286 1.0004. Sp. WOral 0.757 0.464 0.543 1.0005. Sp. Frog 0.226 0.308 0.169 0.208 1.0006. Sp. CELF 0.393 0.171 0.256 0.381 0.081 1.0007. Eng. Comp 0.464 0.314 0.377 0.420 0.048 0.198 1.0008. Eng. Basic 0.526 0.616 0.278 0.373 0.267 0.111 0.650 1.0009. Eng. Peabody 0.234 0.077 0.377 0.299 0.068 0.105 0.582 0.379 1.00010. Eng. WOral 0.305 0.100 0.414 0.365 0.012 0.128 0.807 0.496 0.666 1.00011. Eng. Frog 0.031 − 0.039 0.066 − 0.005 0.025 − 0.035 0.371 0.258 0.318 0.352 1.00012. Eng. CELF 0.221 0.018 0.127 0.210 − 0.111 0.354 0.502 0.227 0.370 0.542 0.208 1.00013. Language 0.419 0.388 0.277 0.306 0.128 0.024 0.546 0.549 0.363 0.501 0.181 0.283 1.00014. Math 0.390 0.303 0.306 0.299 0.126 0.106 0.533 0.388 0.419 0.528 0.147 0.438 0.690 1.00015. Reading 0.471 0.361 0.368 0.414 0.213 0.102 0.615 0.507 0.505 0.579 0.224 0.370 0.820 0.687 1.000Mean 486.941 515.206 59.344 487.008 98.354 11.664 493.483 505.923 85.054 492.518 97.658 11.492 610.414 649.442 641.845SD 11.662 22.748 8.203 10.782 1.485 2.160 10.878 15.711 13.962 12.015 1.388 2.097 27.689 31.811 26.719

Note. Sp.=Spanish; Comp.=Comprehensive Reading; WOral=Woodcock Oral; CELF=Clinical Evaluation of Language Fundamentals; Eng.=English; SD= standarddeviation.

Cambridge Core term

s of use, available at https://ww

w.cam

bridge.org/core/terms. https://doi.org/10.1017/S0267190517000137

Dow

nloaded from https://w

ww

.cambridge.org/core. U

niversity of Maryland - U

niversity Libraries, on 27 Sep 2017 at 02:57:18, subject to the


https://doi.org/10.1017/S0267190517000137



table 4. Global Fit Indices for Measurement Models (CFAs) and Latent VariableRegression Models

Model χ2 df CFI SRMR RMSEA [90% CI] AIC

1. Seven-factor CFA, withacademic achievement inEng. and literacy,school-based language, andnatural language factorswithin both Eng. and Sp.

130.51 67 .95 .12 .070 [.052, .088] 17,444.68

2. Five-factor CFA, withschool-based and naturallanguage combined into anoral language factor withinboth Eng. and Sp.

143.43 78 .95 .121 .066 [.049, .083] 17,435.60

3. Four-factor CFA; same asM2 except with a singleEng. proficiency factor(combining literacy,school-based language, andnatural language)

186.80 82 .92 .16 .082 [.066, .097] 17,470.97

4. Latent regressionpredicting achievementfrom Eng. proficiency, Sp.literacy, and Sp. orallanguage; statisticallyequivalent to M3

186.80 82 .92 .16 .082 [.066, .097] 17,470.97

5. Latent regression nestedwithin M4; path from Sp.oral language toachievement set to 0

188.45 83 .92 .16 .082 [.066, .097] 17,470.617

6. Latent regression nestedwithin M4 and M5, pathsfrom Sp. oral language andSp. literacy to achievementset to 0

188.45 83 .92 .16 .082 [.066, .097] 17,470.617

Note. df = model degrees of freedom; CFI = comparative fit index; SRMR = standardized rootmean square residual; RMSEA = root mean square error of approximation; CI = confidence interval;AIC = Akaike information criterion; Eng. = English; Sp. = Spanish; CFA = confirmatory factoranalysis; M = model.

Spanish M = 98.46, SD = 1.39). Accordingly, this restriction of range limits thecorrelation of the frog story with other measures.

Measurement Models

Alternative measurement models (i.e., confirmatory factor analysis [CFA] models)were fit to the data to examine the hypothesized relationships between the lan-guage and achievement measures; adequate measurement models are a necessaryprecursor to specifying structural relations in a latent regression analysis. Globalfit statistics for measurement and latent variable regression models are presentedin Table 4. First, a seven-factor CFA was specified that included an academic



https://doi.org/10.1017/S0267190517000137



achievement factor and separate factors for literacy, school-based language, andnatural language in English and Spanish underlying the six language measureswithin each language. Construct overlap between corresponding measures in dif-ferent languages is a plausible source of covariance not accounted for adequatelyby the factor structure, so we examined whether it was necessary to allow all re-spective measures to covary between languages. Based on modification indices andstatistical significance of the estimated parameters across different models, onlybetween-language error covariances for basic reading and the CELF were deemednecessary and were included in this and all subsequent models. This seven-factormodel (Model 1) fit the data well, χ2 (67) = 130.51, p < .001, RMSEA = .07,90% CI [.05, .09], CFI = .95, SRMR = .12. All factors were significantly (p <

.001) and positively correlated with academic achievement except Spanish naturallanguage (r = .28, p = .12): English literacy (.72), English school language (.66),English natural language (.78), Spanish literacy (.53), and Spanish school language(.49). Additionally, correlations among language factors within Spanish and withinEnglish were statistically significant and exceeded .85; estimated correlations ofEnglish school-based language with English natural language and English literacywere slightly out of bounds (>1.0) within this model.

To investigate the need for maintaining separate factors for school and naturallanguage measures within each language, a second CFA was specified that includedseparate factors for literacy and oral language development within each language(i.e., school-based and natural language loaded on a single oral language factor).This five-factor model (Model 2) fit the data similarly well, χ2 (78) = 143.43,p < .001, RMSEA = .07, 90% CI [.05, .08], CFI = .95, SRMR = .12. All factorswere significantly correlated with academic achievement (p < .001): English lit-eracy (.71), English oral language (.66), Spanish literacy (.53), and Spanish orallanguage (.47).

Given the very high correlation between English literacy and oral proficiency(.91), as well as concerns about encountering multicollinearity problems in thelatent regression models, one additional CFA was specified in which a single factorof English proficiency underlay the six English language measures. Both literacyand oral language factors were maintained for the Spanish measures given the focalgoal of estimating separate effects for these factors on academic achievement. Thefit of this four-factor model (Model 3) was marginally adequate, χ2 (82) = 186.80,p < .001, RMSEA = .08, 90% CI [.06, .10], CFI = .92, SRMR = .16.

Latent Regression Models

Latent regressions were specified to estimate the unique effects of the variouslanguage factors on academic achievement, reflecting the three theories beingevaluated. As expected, when latent regressions were estimated using multiple-factor structures of both English and Spanish language, multicollinearity resultingfrom high correlations among language and literacy factors within each languagecaused instability in the regression coefficients (even changing signs) and inflatedstandard errors. To minimize these difficulties, the highly correlated English oral



https://doi.org/10.1017/S0267190517000137



figure 2. Standardized Path Coefficients for Latent Regression Models Predicting Aca-demic Achievement. The two sets of partial regression coefficients reflect coefficients forModels 4 and 5, respectively; the path from L1 oral language was constrained to 0 inModel 5.

language and literacy factors were combined into a single English language pro-ficiency factor (as in Model 3) in the latent regression analyses.

Model 4 was estimated to address the threshold hypothesis, in which Spanishlanguage proficiency (Spanish language and literacy) and English languageproficiency (English language and literacy) predict academic achievementmeasured in English. Because all paths were freely estimated, the fit of Model 4was equivalent to its respective measurement model (Model 3). Figure 2 showspartial regression coefficients for Model 4 (first set of estimates); Spanish literacy(β1 = .50) and English language proficiency (β3 = .64) contributed positivelyand significantly to the prediction of academic achievement, whereas Spanishoral language (β2 = −.30) did not contribute to academic achievement aftercontrolling for Spanish literacy and English language proficiency. In the CFA(Model 3), however, Spanish oral language was correlated positively with academicachievement (r = .47), so this negative, nonsignificant estimated coefficient maybe attributable to multicollinearity due to the substantial correlation between



https://doi.org/10.1017/S0267190517000137



Spanish literacy and oral language (r = .88). The three predictors accounted for58% of the variance in the academic achievement factor, which dropped to 56%when Spanish oral language was excluded (Model 5).

Models 5 and 6, respectively, sequentially constrained to 0 the paths from L1oral language (β2) and L1 literacy (β1) to academic achievement; both modelshad similar, marginally adequate fit to the data. To comparatively evaluate both thetransfer theory and the time-on-task theory, the contribution of Spanish literacy toacademic achievement over and above English language proficiency was exam-ined. A chi-square difference test comparing Model 5 to Model 6 was statisticallysignificant, �χ2 (1) = 10.27, p < .01, indicating that eliminating the path fromSpanish literacy to academic achievement significantly worsened the fit of themodel and decreased the R2 from 56% (Model 5) to 52% (Model 6), supportingthe positive contribution of Spanish literacy to the prediction of academic achieve-ment beyond that contributed by English language proficiency. In Model 5 (secondset of coefficients on Figure 2), the standardized partial regression coefficients forSpanish literacy (β1 = .24) and English language proficiency (β3 = .60) were bothstatistically significant.

discussion and conclusions

Our model results indicated that Spanish literacy together with English languageproficiency are substantially predictive of academic achievement in English, con-sistent with the transfer theory. Our examination of transfer theory was basedon evaluating the contribution of Spanish literacy to achievement, controlling forEnglish language proficiency. We noted that this model was a more focused versionof the threshold hypothesis, in that the weaker predictor of Spanish oral languageproficiency was excluded. Model effect sizes were nearly as large as for the morecomplex threshold hypothesis.

Upon a closer look at both the correlations of measures with achievement scoresas well as regression coefficients, we found that the Spanish language sampleand the Spanish CELF vocabulary measure were not substantially correlated withacademic achievement. In comparison, the Spanish literacy and English oral pro-ficiency measures and, to a lesser extent, the more school-oriented Spanish oralproficiency measures (Peabody and Woodcock Oral Language Cluster) were morestrongly positively related to academic achievement.

The time-on-task theory emphasizes L2 (English) language proficiency at thecost of instructional time spent learning in the L1 (Spanish). Our findings didnot support this approach. In particular, Spanish literacy significantly predictedacademic achievement measured in English over and above English languageproficiency (R2 increase of 4% over the 52% of variance accounted for by Englishproficiency). We conducted an additional analysis that included Spanish literacyas a sole predictor of academic achievement and found that it accounted for29% of the academic achievement in English. So, students would be well servedacademically by a language program that supports their growth in literacy in L1while learning L2.



https://doi.org/10.1017/S0267190517000137



A long-standing controversy in language minority education, with importantconsequences for curriculum and instruction for ELLs, is the nature of bilingual-ism and language growth in immigrant children and how their bilingualism mayor may not be related to school success. An outcome of the conception of pro-ficiency embedded within the threshold hypothesis and related notions has beenthe widespread belief among teachers and educational researchers that ELLs mayenter into a state of “semilingualism” (Cummins, 1979) or “limited bilingualism”(Cummins, 1981), in which children are said to know neither their L1 nor their L2with native-like proficiency. As has been pointed out previously (Edelsky, 2006;Edelsky et al., 1983; MacSwan, 2000; MacSwan & Rolstad, 2003, 2006, 2010; Rol-stad, 2015; Wiley, 2005; Wiley & Rolstad, 2014), talk of such linguistic hierarchiesstands to negatively impact children and their communities in unintended ways,and these assertions lack the empirical and theoretical support one should expect.

The transfer theory, however, avoids these pitfalls. This theory, a traditionalnotion in bilingual education policy and embedded in Krashen’s (1996) work aswell as Cummins’s (1981, 2008, 2017) linguistic interdependence hypothesis, sug-gests that academic subject knowledge transfers and indeed facilitates academicdevelopment in the L2 environment. Here, we do not point to differences in thenative language ability of ELLs (and hence to that of their families and communi-ties) as a factor predicting school success, but rather focus on elements indigenousto school culture itself, namely, the learning of literacy and other school-basedsubjects. We prefer to refer to this model as the transfer theory rather than the lin-guistic interdependence hypothesis because we believe that the latter erroneouslyequates language development with academic content knowledge development,while transfer theory (for us) does not.

We have further shown that the time-on-task theory, which seeks to provide arationale for English-only and English-focused programs, is not supported by ourempirical findings, consistent with program evaluation reviews (McField & Mc-Field, 2014; Reljic et al., 2015; Rolstad et al., 2005; Slavin & Cheung, 2005). Thecore conceptual problem with the time-on-task theory is its failure to distinguishbetween academic content learning, independent of a linguistic medium, and thelinguistic medium itself. Once this distinction is properly understood, the time-on-task principle—that “the amount of time spent learning a subject is the greatestpredictor of achievement in that subject” (Rossell & Baker, 1996, p. 22)—is en-tirely consistent with the transfer theory, and is supportive of bilingual education:The amount of time spent learning academic subjects may be the greatest predictorof achievement in those subjects, but time spent learning will be meaningless if itis not spent in a language children can understand.

limitations

We note two limitations of our study. First, we were not able to track student pro-gram placements in English-only and bilingual education programs consistentlydue to inadequate documentation and unanticipated student program changes. We



https://doi.org/10.1017/S0267190517000137



addressed this limitation by focusing on the hypothesized intermediate effects ofdifferent program types noted in Table 1. It would be interesting in future studiesto examine relationships between particular program features and academic out-comes. Second, our student participants were drawn from schools across sevenschool districts. Accordingly, the data would be regarded as clustered to somedegree due to multiple students being sampled from particular schools that werenested within these districts. Although this violates the assumption of indepen-dence in sampling, we did not have sufficient numbers of school or district clustersto account statistically for this nonindependence in the data structure. It is possiblethat the standard errors of our parameter estimates are slightly smaller than mightbe expected if data were sampled randomly, which can increase the chance offinding significant effects.

implications

Theories are important because they provide a conceptual framework from whicha wide range of inferences may be drawn. Our present study sought to gain insightinto competing theories of academic achievement differences for bilingual studentsprecisely because different theories hold very different sets of implications foreducational policy and practice. For instance, the threshold hypothesis may implyfor some that instruction should focus specifically on language-related aspectsof L1 development, attending to vocabulary and grammar in the L1 rather thanusing the L1 as a medium for teaching and learning in content areas. Further, thethreshold hypothesis may suggest for some that L1 development needs to precedeL2 teaching, with the goal of achieving a developmental threshold in the L1 beforethe L2 can effectively become a focus of instruction. The threshold hypothesispresupposes that the language children bring with them to school, and hence thelanguage of their community, is hierarchically related to the language of schoolingand is in need of further development to facilitate school success. The time-on-tasktheory, on the other hand, which stems from an entirely different understanding ofthe basic descriptive facts, implies that teachers of ELLs should devote as muchtime as possible to English-mediated instruction, making little or no use of the L1as a medium of instruction.

Transfer theory, in contrast to both the threshold hypothesis and the time-on-task theory, conceptualizes the L1 as a vehicle by which content knowledge isconveyed to learners and is not itself in need of improvement or development.Literacy is a domain of language use specific to the school setting and does notrepresent a further stage of linguistic development, as is commonly believed. Thisperspective views school as a specific context for language use and does not treatschool language as a special language variety corresponding to special cognitiveabilities that are inaccessible to language users in out-of-school contexts (MacSwan& Rolstad, 2003, 2005, 2010; Rolstad, 2015). Development of children’s home lan-guage is not a prerequisite to English language learning, so the teaching of ESL mayproceed concurrently and independently. Thus, bilingual education contributes to



https://doi.org/10.1017/S0267190517000137



children’s success at school because it provides them with access to content areaknowledge so they can keep up academically during the time it takes them to learnEnglish.

The concept of second language instructional competence (SLIC) serves asa complement to transfer theory (MacSwan & Rolstad, 2003; Rolstad, 2005,2015; Rolstad & MacSwan, 2008). Offered as an alternative to Cummins’s BasicInterpersonal Communication Skills/Cognitive-Academic Language Proficiency(BICS/CALP) dichotomy, SLIC refers to children’s development of L2 profi-ciency, not to proficiency in their home or community language; once childrenhave learned English well enough to understand school subject matter instructionin English, they have developed SLIC. Second language acquisition is also facili-tated as a result of L1 instruction because, as Krashen (1996) argued, it strengthenschildren’s knowledge of the context for L2 acquisition, contributing to the com-prehensibility of L2 input (on the role of background knowledge in L2 acquisition,see Bygate, 2001; Gass & Varonis, 1984). Thus, transfer theory views children’shome language as a critically important resource that allows access to compre-hensible content area instruction mediated through the home language with theconcurrent teaching of ESL. As children’s knowledge of school content grows, itpromotes the acquisition of English by creating a more effective context for L2teaching and learning, with the goal of gradually achieving SLIC when childrencan meaningfully and effectively engage with English-mediated instruction.

From a policy perspective, transfer theory suggests that children’s home lan-guage should be actively used as a resource for the purposes of teaching andlearning. It contrasts with the threshold hypothesis, which posits a developmentalrelationship between the home language and the L2, and with the time-on-task the-ory, which emphasizes English-mediated instruction and views use of children’shome language in classroom settings as an essential waste of instructional timeand effort. Transfer theory views children’s emerging bilingualism holistically(Grosjean, 1985, 2010), permitting and encouraging access to their full linguisticrepertoire (García, 2009; MacSwan, 2017) as an ongoing resource to aid theirsuccess in school.

Acknowledgments

We gratefully acknowledge research support from the U.S. Department of Edu-cation Institute of Education Science, award # R305T000164. We are grateful forthe support and assistance of students, teachers, administrators, and undergraduateand graduate student research assistants, especially Jessie Ortiz.

notes1. English-only programs continue to predominate in school systems across the United States despiteresearch clearly favoring first language support for a variety of political and practical reasons. SeeCrawford (2004, 2008) for discussion.2. For a detailed description of program models, see Baker and Wright (2017).



https://doi.org/10.1017/S0267190517000137



referencesAugust, D., & Hakuta, K. (Eds.). (1997). Educating language-minority children. Washington, DC:

National Academy Press.August, D., & Shanahan, T. (Eds.). (2006). Developing literacy in second-language learners: Report

of the National Literacy Panel on Language-Minority Children and Youth. Mahwah, NJ: Erlbaum.Baker, C., & Wright, W. (2017). Foundations of bilingual education and bilingualism (6th ed.). Bristol,

UK: Multilingual Matters.Baker, K., & de Kanter, A. A. (1981). Effectiveness of bilingual education: A review of the literature

(Final Draft Report). Washington, DC: Department of Education Office of Planning, Budget, andEvaluation.

Brown, R. (1973). A first language. Cambridge, MA: Harvard University Press.Bygate, M. (2001). Effects of task repetition on the structure and control of language. In M. Bygate,

P. Skehan, & M. Swain (Eds.), Task-based learning: Language teaching, learning and assessment(pp. 23–48). London, UK: Longman.

Chomsky, N. (1995). The minimalist program. Cambridge, MA: MIT Press.Commins, N., & Miramontes, O. (1989). Perceived and actual linguistic competence: A descriptive

study of four low-achieving Hispanic bilingual students. American Educational Research Journal,26(4), 443–472.

Cook, V. J. (2001). Using the first language in the classroom. Canadian Modern Language Review,57(3), 402–423.

Crawford, J. (2004). Educating English learners: Language diversity in the classroom (5th ed.). LosAngeles, CA: Bilingual Education Services.

Crawford, J. (2008). Advocating for English learners. Bristol, UK: Multilingual Matters.Cummins, J. (1976). The influence of bilingualism on cognitive growth: A synthesis of research findings

and explanatory hypotheses. Working Papers in Bilingualism, 9, 1–43.Cummins, J. (1979). Linguistic interdependence and the educational development of bilingual children.

Review of Educational Research, 49, 221–251.Cummins, J. (1981). The role of primary language development in promoting educational success

for language minority students. In California State Department of Education (Ed.), Schooling andlanguage minority students: A theoretical framework. Sacramento, CA: Office of Bilingual BiculturalEducation, California State Department of Education.

Cummins, J. (1994). Semilingualism. In R.R. Asher (Ed.), International encyclopedia of language andlinguistics (2nd ed., pp. 3812–3814). Oxford, UK: Elsevier Science.

Cummins, J. (2000). Language, power, and pedagogy: Bilingual children in the crossfire. Clevedon,UK: Multilingual Matters.

Cummins, J. (2008). Teaching for transfer: Challenging the two solitudes assumption in bilingualeducation. In J. Cummins & N. H. Hornberger (Eds.), Encyclopedia of language and education:Vol. 5. Bilingual education (2nd ed., pp. 65–75). Boston, MA: Springer Science + BusinessMedia.

Cummins, J. (2017). Teaching for transfer in multilingual educational contexts. In O. Garcia & A. Lin(Eds.), Encyclopedia of language and education: Vol. 5. Bilingual education (3rd ed., pp. 103–115).New York, NY: Springer Science + Business Media.

Curtiss, C. (2013). Revisiting modularity: Using language as a window to the mind. In M. Piatelli-Palmarini & R. C. Berwick (Eds.), Rich languages from poor inputs (pp. 68–90). Oxford, UK: OxfordUniversity Press.

Curtiss, S., MacSwan, J., Schaeffer, J., Kural, M., & Sano, T. (2004). GCS: A grammatical codingsystem for natural language data. Behavior Research Methods, Instruments, and Computers, 34(3),459–480.

Edelsky, C. (2006). With literacy and justice for all: Rethinking the social in language and education(3rd ed.). New York, NY: Routledge.

Edelsky, C., Hudelson, S., Flores, B., Barkin, F., Altweger, J., & Jilbert, K. (1983). Semilingualismand language deficit. Applied Linguistics, 4, 1–22.

Faltis, C. (1989). Code-switching and bilingual schooling: An examination of Jacobson’s new concur-rent approach. Journal of Multilingual and Multicultural Development, 10(2), 117–127.

Gallistel, C. R., & King, A. P. (2009). Chapter 13: The modularity of learning. In Memory and the com-putational brain: Why cognitive science will transform neuroscience. Oxford, UK: Wiley-Blackwell.doi:10.1002/9781444310498.ch13

García, O. (2009). Bilingual education in the 21st century: A global perspective. Malden, MA:Basil/Blackwell.


http://dx.doi.org/10.1002/9781444310498.ch13


https://doi.org/10.1017/S0267190517000137



Gass, S. M., & Varonis, M. (1984). The effect of familiarity on the comprehensibility of non-nativespeech. Language Learning, 34(1), 65–89.

Genesee, F., Lindholm-Leary, K., Saunders, W. M., & Christian, D. (Eds.). (2006). Educating Englishlanguage learners: A synthesis of research evidence. New York, NY: Cambridge University Press.

Greene, J. P. (1998). A meta-analysis of the effectiveness of bilingual education. Claremont, CA:Thomas Rivera Policy Institute.

Grosjean, F. (1985). The bilingual as a competent but specific speaker-hearer. Journal of Multilingualand Multicultural Development, 6(6), 467–477.

Grosjean, F. (2010). Bilingual: Life and reality. Cambridge, MA: Harvard University Press.Hakuta, K., & D’Andrea, D. (1992). Some properties of bilingual maintenance and loss in Mexican

background high-school students. Applied Linguistics, 13(1), 72–99.Hansegård, N. E. (1968). Tvåspråkighet eller halvspråkighet? Stockholm: Aldus series 253.Jacobson, R. (1978). Code-switching in south Texas: Sociolinguistic considerations and pedagogical

applications. Journal of the Linguistic Association of the Southwest, 3, 20–32.Jacobson, R. (1990). Allocating two languages as a key feature of bilingual methodology. In R. Jacobson

& C. Faltis (Eds.), Language distribution issues in bilingual schooling (pp. 3–17). Clevedon, England:Multilingual Matters.

Kena, G., Hussar, W., McFarland, J., de Brey, C., Musu-Gillette, L., Wang, X., … Dunlop Velez, E.(2016). The condition of education 2016 (NCES 2016–144). Washington, DC: U.S. Department ofEducation, National Center for Education Statistics.

Krashen, S. (1996). Under attack: The case against bilingual education. Culver City, CA: LanguageEducation Associates.

Lau v. Nichols, 414 U.S. 563 (1974).MacSwan, J. (2000). The threshold hypothesis, semilingualism, and other contributions to a deficit

view of linguistic minorities. Hispanic Journal of Behavioral Sciences, 22(1), 3–45.MacSwan, J. (2017). A multilingual perspective on translanguaging. American Educational Research

Journal, 54(1), 167–201. doi:10.3102/0002831216683935MacSwan, J., & Pray, L. (2005). Learning English bilingually: Age of onset of exposure and rate of

acquisition of English among children in a bilingual education program. Bilingual Research Journal,29(3), 687–712.

MacSwan, J., & Rolstad, K. (2003). Linguistic diversity, schooling, and social class: Rethinking ourconception of language proficiency in language minority education. In C. B. Paulston & R. Tucker(Eds.), Sociolinguistics: The essential readings (pp. 329–340). Oxford, UK: Blackwell.

MacSwan, J., & Rolstad, K. (2005). Modularity and the facilitation effect: Psychological mecha-nisms of transfer in bilingual students. Hispanic Journal of the Behavioral Sciences, 27(2), 224–243.

MacSwan, J., & Rolstad, K. (2006). How language tests mislead us about children’s abilities: Impli-cations for special education placements. Teachers College Record, 108(11), 2304–2328.

MacSwan, J., & Rolstad, K. (2010). The role of language in theories of academic failure for linguisticminorities. In J. Petrovic (Ed.), International perspectives on bilingual education: Policy, practice,and controversy (pp. 173–195). Charlotte, NC: Information Age.

MacWhinney, B. (1995). The CHILDES project: Tools for analyzing talk (2nd ed.). Mahwah, NJ:Erlbaum.

Martin-Jones, M., & Romaine, S. (1986). Semilingualism: A half-baked theory of communicativecompetence. Applied Linguistics, 7(1), 26–38.

Mayer, M. (1969). Frog, where are you? New York, NY: Puffin.McField, G. P., & McField, D. R. (2014). The consistent outcome of bilingual education programs:

A meta-analysis of meta-analyses. In G. P. McField (Ed.), The miseducation of English learners(pp. 267–298). Charlotte, NC: Information Age.

Meyer, M. M., & Fienberg, S. E. (Eds.). (1992). Assessing evaluation studies: The case of bilingualeducation strategies. Washington, DC: National Academic Press.

Milk, R. (1986). The issue of language separation in bilingual methodology. In E. García & B. Flores(Eds.), Language and literacy research in bilingual education (pp. 67–86). Tempe, AZ: Center forBilingual Education, Arizona State University.

National Academies of Sciences, Engineering, and Medicine. (2017). Promoting the educationalsuccess of children and youth learning English: Promising futures. Washington, DC: NationalAcademies Press.

Paulston, C. B. (1983). Swedish research and debate about bilingualism. Stockholm, Sweden: NationalSwedish Board of Education.


http://dx.doi.org/10.3102/0002831216683935


https://doi.org/10.1017/S0267190517000137



Petrovic, J. E., & Olmstead, S. (2001). Language, power, and pedagogy: Bilingual children in thecrossfire, by J. Cummins [Book review]. Bilingual Research Journal, 25(3), 405–410.

Porter, R. P. (1990). Forked tongue. The politics of bilingual education. New York, NY: Basic Books.Reljic, G., Ferring, D., & Martin, R. (2015). A meta-analysis on the effectiveness of bilingual programs

in Europe. Review of Educational Research, 85(1), 92–128.Ringbom, H. (1962). Tvåspråkighet som forskningsobjekt. Finsk Tidskrift, 6.Rolstad, K. (2005). Rethinking academic language in second language instruction. In J. Cohen,

K. McAlister, K. Rolstad, & J. MacSwan (Eds.), ISB4: Proceedings of the 4th International Sympo-sium on Bilingualism (pp. 1993–1999). Somerville, MA: Cascadilla Press.

Rolstad, K. (2015). Second language instructional competence. International Journal of BilingualEducation and Bilingualism, 1, 1–13. doi:10.1080/13670050.2015.1057101

Rolstad, K., & MacSwan, J. (2008). BICS/CALP: Theory and critique. In J. Gonzalez (Ed.), Encyclo-pedia of bilingual education (pp. 62–65). Thousand Oaks, CA: Sage.

Rolstad, K., Mahoney, K., & Glass, G. V. (2005). The big picture: A meta-analysis of program effec-tiveness research on English language learners. Educational Policy, 19, 572–594.

Rossell, C. H., & Baker, K. (1996). The educational effectiveness of bilingual education. Research inthe Teaching of English, 30(1), 7–74.

Semel, E., Wiig, E. H., & Secord, W. A. (1995). Clinical Evaluation of Language Fundamentals R –Third Edition (CELF R ). San Antonio, TX: The Psychological Corporation.

Skutnabb-Kangas, T. (1981). Bilingualism or not? The education of minorities. (L. Malmberg & D.Crane, Trans.). Clevedon, UK: Multilingual Matters.

Slavin, R., & Cheung, A. (2005). A synthesis of research of reading instruction for English languagelearners. Review of Educational Research, 75(2), 247–284.

Thompson, M. S., DiCerbo, K., Mahoney, K. S., & MacSwan, J. (2002). ¿Éxito en California? A validitycritique of language program evaluations and analysis of English learner test scores. Education PolicyAnalysis Archives, 10(7). doi:10.14507/epaa.v10n7.2002

Toukomaa, P., & Skutnabb-Kangas, T. (1977). The intensive teaching of the mother tongue to migrantchildren at pre-school age (Department of Sociology and Social Psychology Research Report No.26). Tampere, Finland: University of Tampere.

United States Commission on Civil Rights. (1975). A better chance to learn: Bilingual biculturaleducation. Washington, DC: Government Printing Office.

Uriagereka, J. (2012). Spell-out and the minimalist program. Oxford, UK: Oxford University Press.Valadez, C., MacSwan, J., & Martínez, C. (2000). Toward a new view of low-achieving bilinguals: A

study of linguistic competence in designated “semilinguals.” Bilingual Review, 25(3), 238–248.Wiley, T. G. (2005). Literacy and language diversity in the United States (2nd ed.). Washington, DC:

Center for Applied Linguistics.Wiley, T. G., & K. Rolstad. (2014). The common core state standards and the great divide. International

Multilingual Research Journal, 8(1), 38–55.Willig, A. C. (1985). A meta-analysis of selected studies on the effectiveness of bilingual education.

Review of Educational Research, 55(3), 269–318.Wong Fillmore, L., & Valadez, C. (1986). Teaching bilingual learners. In M. C. Wittrock (Ed.), Hand-

book of research on teaching (3rd ed.). New York, NY: Macmillan.Woodcock, R. W. (1991). Woodcock language proficiency battery (revised). Itasca, IL: Riverside.


http://dx.doi.org/10.1080/13670050.2015.1057101

http://dx.doi.org/10.14507/epaa.v10n7.2002


https://doi.org/10.1017/S0267190517000137


three theories of the effects of language education programs: …macswan/macswanetal2017.pdf ·...

Documents